Video series · 6 parts
Short lessons on building business automations that hold up: knowing what you automate, granularity, incremental rollout, reliability, logging and human approval.

You can't reliably automate a task you haven't first standardized and done manually. I explain why every automation needs written instructions or an SOP, and why I always do the task by hand at least once or twice…

More granularity in how you describe a process is what makes AI automations reliable, because if you can’t explain every tiny step in detail, you probably don’t understand the process well enough to automate it. I show…

Automation isn't an all-or-nothing process: I argue that automating 80% of a task beats automating none of it, and I show how that plays out in things like video editing and social media management—if you spend 10…

I argue that most AI agents are hype and that reliable automations come from deterministic code, with LLMs used only where actual reasoning is required. I show how I build fast, cheap automations by writing the whole…

I argue that proper detailed logs and immediate alerts are the difference between catching a broken AI automation in minutes and discovering it weeks later. I show why every automation step needs to be logged…

Always require a human approval step before an AI agent runs any sensitive or irreversible action. I point to real incidents where autonomous agents caused catastrophic damage: one AI agent deleted a company's database…


Dude, you’re at the bottom of my landing page.
Just book a call already if you’re that interested.
You scrolled all the way here.